US20060044013A1 - Hybrid CVSL pass-gate level-converting sequential circuit for multi-Vcc microprocessors - Google Patents
Hybrid CVSL pass-gate level-converting sequential circuit for multi-Vcc microprocessors Download PDFInfo
- Publication number
- US20060044013A1 US20060044013A1 US10/924,906 US92490604A US2006044013A1 US 20060044013 A1 US20060044013 A1 US 20060044013A1 US 92490604 A US92490604 A US 92490604A US 2006044013 A1 US2006044013 A1 US 2006044013A1
- Authority
- US
- United States
- Prior art keywords
- stage storage
- circuit
- data
- inverter
- stage
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Classifications
-
- H—ELECTRICITY
- H03—ELECTRONIC CIRCUITRY
- H03K—PULSE TECHNIQUE
- H03K3/00—Circuits for generating electric pulses; Monostable, bistable or multistable circuits
- H03K3/02—Generators characterised by the type of circuit or by the means used for producing pulses
- H03K3/353—Generators characterised by the type of circuit or by the means used for producing pulses by the use, as active elements, of field-effect transistors with internal or external positive feedback
- H03K3/356—Bistable circuits
- H03K3/356104—Bistable circuits using complementary field-effect transistors
- H03K3/356113—Bistable circuits using complementary field-effect transistors using additional transistors in the input circuit
- H03K3/356147—Bistable circuits using complementary field-effect transistors using additional transistors in the input circuit using pass gates
- H03K3/356156—Bistable circuits using complementary field-effect transistors using additional transistors in the input circuit using pass gates with synchronous operation
Definitions
- This invention generally relates in one or more of its embodiments to electronic circuits which perform a level-conversion function embedded in a sequencing element.
- FIG. 1 is a diagram showing a level-conversion function that may be performed in accordance with at least one embodiment of the present invention.
- FIG. 2 is a diagram showing a logic circuit which performs a level-conversion function in accordance with one embodiment of the present invention.
- FIG. 3A is a diagram showing that the keeper circuit in the master stage of FIG. 2 may be formed from one or more clocked inverters
- FIG. 3B is a diagram showing a transistor implementation of the inverter of FIG. 3A
- FIG. 3C is a diagram showing that the keeper circuit in the slave stage of FIG. 2 may be formed from clocked inverters
- FIG. 3D is a diagram showing another circuit arrangement for the slave-stage keeper circuit.
- FIG. 4 is a flow chart showing functional blocks included in a method for performing a level conversion in accordance with one embodiment of the present invention.
- FIG. 5 is a diagram showing a logic circuit which performs a level-conversion function in accordance with another embodiment of the present invention.
- FIG. 6A shows an alternative arrangement for the master-stage keeper circuit
- FIG. 6B shows an alternative implementation for one or more of the inverters which form the slave-stage keeper circuit.
- FIG. 7 is a flow chart showing functional blocks included in a method for performing a level conversion in accordance with another embodiment of the present invention.
- FIG. 8 is a diagram showing a processing system which may include a logic circuit and/or perform a level-conversion function in accordance with one or more embodiments of the present invention.
- FIG. 1 shows a logic circuit which performs a level-conversion function in accordance with one embodiment of the present invention.
- a first voltage signal 1 is converted into a second voltage signal 2 .
- the first voltage signal (V cc1 ) is shown as having a low voltage and the second voltage signal (V cc2 ) is shown as having a high voltage.
- the opposite conversion may be performed, e.g., a high-voltage signal may be converted into a low-voltage signal.
- level-conversion circuitry is preferably embedded in the logic circuit in order to reduce the size and overhead of the host system while simultaneously increasing its functionality and throughput.
- the logic circuit is shown to be a D flip-flop driven by a clock signal 3 which depending on the embodiment may be the same as or generated based on either or only one of V cc1 or V cc2 .
- the clock signal may be different and/or other types of flip-flops, latches, and/or logic circuits may be adapted to perform the level-conversion function of the one or more embodiments of the present invention described herein.
- the type and/or magnitude of the level conversion to be performed may be adapted to meet the requirements of an intended application.
- FIG. 2 shows a circuit diagram of a D flip-flop having an embedded level-conversion circuit in accordance with one embodiment of the present invention.
- This flip-flop includes a master stage 10 and a slave stage 20 operating in different voltage supply domains.
- the master stage operates in a first supply domain and includes a pass gate 11 , an NMOS pull-down transistor 12 , a keeper circuit 13 , and an NMOS pull-down transistor 14 .
- the keeper circuit is preferably formed from two cross-coupled inverters 15 and 16 and includes a storage node SN 1 for storing a data value input through the pass gate and a feedback node FN 1 for storing the complement of the input data value.
- the storage node is coupled to the gate of transistor 12 and the feedback node is coupled to the gate of transistor 14 .
- the slave stage operates in a second supply domain and includes transistors 21 and 22 and a keeper circuit 23 .
- the keeper circuit may be formed from two cross-coupled inverters 24 and 25 and includes a storage node SN 2 for storing the same logical value stored at node SN 1 .
- Feedback node FN 2 stores the complement of the value stored at SN 2 .
- the logical values stored at SN 1 and SN 2 are the same (e.g., logical 0 or 1), but the actual voltages corresponding to these values may differ by virtue of the different power supplies driving the master and slave stages.
- Storage node SN 2 is coupled to a terminal (drain or source) of transistor 21 and feedback node FN 2 is coupled to a terminal (drain or source) of transistor 22 .
- the gates of these transistors are coupled to a clock signal used to alternately activate the master and slave stages.
- the master and slave stages operate in different domains by virtue of the different power supplies coupled to the keeper circuits.
- the master stage thus operates at low voltage and the slave stage performs the level conversion.
- the different supplies effect the conversion of the low-voltage data value stored in the master stage keeper circuit into a high-voltage data value stored in the slave stage keeper circuit.
- the values of V ccl and V cch therefore control the level-conversion that takes place at the output of the flip-flop. (In an alternative embodiment, the opposite level-conversion may be performed).
- the low and high voltage supplies are also preferably coupled to input inverter 16 and output inverter 36 .
- V cc1 may be less than, equal to, or greater than V cc2 and clock signal clk can be equal to or operate based on either V cc1 or V cc2 .
- the clock signal may only operate on V cc2 .
- the logical values stored in keeper circuits 13 and 23 are the same (e.g., 0 or 1), but the analog voltages that represent these logical values may be different, e.g., a logical 1 in circuit 13 may be equal to or based on V cc1 and a logical 1 in circuit 23 may be equal to or based on V cc2 .
- clk and clkd are logically the same value but “d” is added to indicate a delayed clock signal.
- clock signal clk activates the master and slave stages of the flip-flop at different times.
- the clock signal is passed through two inverters 41 and 42 prior to being input into the master stage. These inverters may be driven by the slave-stage power supply (e.g., V cch ), although a different supply may be used.
- V cch the slave-stage power supply
- the depicted arrangement is preferable for purposes of improving time-borrowing from the master stage to the slave stage. This, in turn, improves the minimum D-Q path delay of the flip-flop.
- the clock signal may be delayed using a different circuit arrangement.
- a high value (e.g., logical 1) is output from inverter 41 and a low value is output from inverter 42 .
- These values activate pass gate 11 , thereby causing the data value from the D input to pass to storage node SN 1 of the keeper circuit 13 .
- inverter 16 is driven by V ccl , the data value transferred through pass gate 11 operates at or based on V ccl .
- the voltage supply used to drive inverters 41 and 42 preferably correspond to high voltage supply V cch , however a different voltage supply may alternatively be used.
- the data value is inverted by inverter 15 .
- the complement of the SN 1 value is stored at FN 1 .
- the values stored at SN 1 and FN 1 are in the low-voltage domain as defined by voltage supply V ccl , which not only drives inverter 16 but which also preferably drives the inverters of the keeper circuit.
- the low value of the clk turns off transistors 21 and 22 , thereby ensuring that the voltages at SN 1 and FN 1 are maintained until a subsequent clock signal transition takes place.
- transistor 12 couples the value stored at SN 1 to the slave stage. So when transistor 21 opens, if the value is 1 then SN 2 will discharge. Conversely, if the value is 0 at SN 1 , when 21 opens nothing will happen to SN 2 , but FN 2 will discharge since transistor 22 will be on and FN 1 would have been 1 turning on transistor 14 .
- a low value (e.g., logical 0) is output from inverter 41 and a high value from inverter 42 .
- a high value of clk turns on transistors 21 and 22 , thereby causing the data values stored in the master-stage keeper to pass to the slave-stage keeper circuit, e.g., the logical value at SN 1 passes to SN 2 and the logical value at FN 1 passes to FN 2 .
- the value at FN 2 then passes to the Q output of the flip-flop through inverter 36 , which is also preferably driven by the slave-stage power supply.
- the slave stage of the flip-flop therefore performs the internal level-conversion function of the flip-flop, converting the input data value from the low-voltage domain (V ccl ) to the high-voltage domain (V cch ) at the output. This is accomplished by driving at least the keeper circuit of the slave stage with a higher voltage supply than is used to drive the master stage keeper. Also, the slave stage inverter 36 is preferably driven using the higher domain voltage and the clock signal clk may be the same as or different from the high-V cc Q output.
- FIG. 2 embodiment performs a level conversion from a low voltage supply domain to a high voltage supply domain
- the opposite level conversion can be performed by driving inverters 15 , 16 , and 17 with V cch and inverters 23 , 24 , 25 with V ccl .
- FIGS. 3A-3D show alternative arrangements for the keeper circuits.
- FIG. 3A shows that inverter 16 may be a clocked inverter which store and output data based on clock signal clk and its complement ⁇ overscore (clk) ⁇ .
- FIG. 3B shows one type of transistor implementation of the inverter of FIG. 3A .
- This implementation includes two PMOS transistors 43 and 44 and two NNOS transistors 45 and 46 coupled between a power supply and reference potential.
- Storage node SN 1 is coupled between transistors 44 and 45
- feedback node FN 1 is coupled to the gates of transistors 43 and 44 .
- the gates of transistors 44 and 45 are coupled to ⁇ overscore (clk) ⁇ and clk respectively.
- FIG. 3C shows that keeper circuit 20 may be formed as an arrangement of clocked inverters, e.g., inverters 23 and 24 store and output data in accordance with clock signal clk.
- FIG. 3D shows another implementation of keeper circuit 20 .
- This implementation includes a cross-coupled arrangement of PMOS transistors 47 through 50 .
- the gates of transistors 48 and 50 are coupled to an external signal, which may be clk. Any combination of the circuits shown in FIGS. 3A-3D may be incorporated into FIG. 2 for purposes of storing data therein.
- FIG. 4 is a flow chart showing functional blocks included in a method for performing a level conversion in accordance with one embodiment of the present invention.
- the method may be performed by the circuit of FIG. 2 or by another circuit.
- the method includes storing an input data value in a first stage of a flip-flop circuit (Block 51 ) and then driving the first stage with a first supply voltage (Block 52 ).
- a second stage of the flip-flop circuit is then activated (Block 53 ), and then the value stored in the first stage of the flip-flop circuit is transferred for storage in the second stage (Block 54 ).
- the second stage is then driven with a second supply voltage different from the first supply voltage, thereby achieving the level conversion.
- the value stored in the second stage is then used as a basis for forming the output of the flip-flop.
- the first and second stages are alternately activated by different states of a clock signal.
- FIG. 5 shows a circuit diagram of a D flip-flop having an embedded level-conversion circuit in accordance with another embodiment of the present invention.
- This flip-flop includes a master stage 100 and a slave stage 110 which operate in a same voltage supply domain.
- An inverter 95 coupled to the D input is driven by a different supply voltage.
- the master and slave stages are shown as being driven by a high voltage supply and inverter 95 by a low voltage supply. This embodiment may prove to be preferable for some applications, since by driving the master stage with a higher voltage supply will make the flip-flop less susceptible to noise.
- the master stage includes a pull-down clocked NMOS transistor 111 , a pull-down clocked NMOS transistor 112 , and a keeper circuit 113 formed from cross-coupled inverters 114 and 115 .
- Transistor 111 is coupled to the storage node SN 1 ′ of the keeper circuit and transistor 112 is coupled to the feedback node FN 1 ′.
- Inverter 95 which inverts the input data value for storage at the feedback node FN 1 ′, is coupled between the gates of two transistors 101 and 102 .
- Transistors 101 and 102 perform pull-down functions. When transistors 111 and 112 turn on, either 101 or 102 will pass a 0 to either SN 1 ′ or FN 1 ′ respectively. If SN 1 ′ is 0 , then inverter 114 will pull FN 1 ′ to 1.
- the slave stage includes a keeper circuit 121 having a storage node SN 2 ′ and feedback node FN 2 ′ disposed between cross-coupled inverters 125 and 126 .
- the storage node of the slave stage is coupled to the feedback node of the master stage through a pass gate 130 , which is operated based on an input clock signal clk 2 .
- a pass gate 130 which is operated based on an input clock signal clk 2 .
- only one clocked inverter is used in order to reduce the clock power of the flip-flop.
- the master and slave stages operate in a same voltage domain (e.g., V cch ) and only the input inverter 95 is driven by a different voltage supply.
- the feedback inverters in the keeper circuits may be sized to be weak or they can be fully interrupted based on the clock signal.
- clock signal clk′ activates the master and slave stages of the flip-flop at different times.
- clk′ transitions from a high value to a low value, a high value is output from inverter 140 , thereby turning on transistors 111 and 112 .
- the data value at the D input and its complement (generated by the output of inverter 95 ) respectively pass to storage and feedback nodes SN 1 ′ and FN 1 ′ of the master keeper circuit. Because the clock is low, the pass gate is not activated and thus the value stored at FN 1 ′ does not pass to the slave stage.
- both inverters 120 and 140 are driven by the same supply voltage, which, for example, may be the same supply voltage used to drive the master and slave stages, e.g., V cch . In an alternative embodiment, these inverters may be driven by different supply voltages, at least one of which may differ from the master and slave stage supplies.
- inverter 150 In the slave stage, the logical value output from inverter 120 is stored at node SN 2 ′ of keeper circuit 121 and the complement of this value is stored at node FN 2 ′. The logical value stored at node SN 2 ′ is passed through an inverter 150 to form the Q output of the flip-flop. To maintain the level conversion, inverter 150 is driven by master- and slave-stage supply voltage, e.g., V cch . Clock signal clk 2 may equal or be based on V cc2 in FIG. 1 , which in this case may be the high-voltage power supply V cch .
- V cch master- and slave-stage supply voltage
- FIG. 6A shows an alternative arrangement for the master-stage keeper circuit.
- inverters 114 and 115 are clocked inverters, which store data based on when clk 2 is high and ⁇ overscore (clk) ⁇ 2 is low.
- FIG. 6B shows an alternative implementation for one or more of the inverters which form the slave-stage keeper circuit.
- inverter 125 is a clocked inverter and the other inverter in this keeper may have a similar arrangement.
- Inverter 125 functions as an inverter when clk 2 is low and ⁇ overscore (clk) ⁇ 2 is high.
- clk 2 is high and ⁇ overscore (clk) ⁇ 2 is low
- the inverter is tri-stated allowing pass gate 30 to write into SN 2 ′ without contention.
- this pass gate is closed, inverter 125 functions like an inverter and when 20 is open the inverter is tri-stated.
- FIG. 7 is a flow chart showing functional blocks included in a method for performing a level conversion in accordance with another embodiment of the present invention.
- the method may be performed by the circuit of FIG. 5 or by another circuit.
- the method includes passing an input data value through an inverter (Block 151 ) and then driving the inverter with a first supply voltage (Block 152 ).
- the inverted value is then transferred for storage into a first stage of a flip-flop circuit. (Block 153 ).
- a second stage of the flip-flop circuit is then activated (Block 154 ), and the value stored in the first stage is transferred through another inverter for storage in the second stage (Block 155 ).
- the first and second stages are driven with at least a second supply voltage.
- the first and second stages may be driven by different supply voltages, both of which are different from the first supply voltage.
- the value stored in the second stage is then used as a basis for forming the output of the flip-flop. (Block 156 ).
- the second stage output value may be inverted.
- the first and second stages are alternately activated by different states of a clock signal.
- level conversion performed by foregoing embodiments are illustratively described as being included in flip-flop circuits.
- the use of flip-flop circuits, while preferable, is by no means a necessity. Rather, the level conversions described herein may be incorporated in a variety of other logic circuits including but not limited to latches, flip-flops, and mux latches.
- the logic circuits may themselves be included in a variety of higher-level circuits or systems, one example of which includes a microprocessor system operating in two or more voltage supply (V cc ) domains.
- V cc voltage supply
- one domain operates based on a low voltage supply (V ccl ) and another domain operates based on a high voltage supply (V cch ).
- V ccl low voltage supply
- V cch high voltage supply
- the first voltage domain can be less than or greater than the second voltage domain, or the domains may be equal, at any time during operation.
- One or more embodiments of the present invention may be used to perform a level-conversion between these domains.
- a logic circuit in accordance with any one or more of the embodiments previously described may be used, for example, to latch or store a data value while simultaneously serving as a level-conversion interface between low-V cc and high-Vcc domains.
- Other higher-level systems which may include a logic circuit in accordance with any of the embodiments described herein include but are not limited to multi-core processors, system-on-chip, core-to-cache interface, and multi-voltage ASICS.
- FIG. 8 is a diagram of a processing system which includes any one or more embodiments of the present invention.
- the processing system includes a processor 210 , a power supply 220 , and a memory 230 which, for example, may be a random-access memory.
- the processor may include an arithmetic logic unit 212 and an internal cache 214 .
- the system may also include a graphical interface 240 , a chipset 250 , a cache 260 and a network interface 270 .
- the processor may be a microprocessor or any other type of processor.
- the processor may be included on a chip die with all or any combination of the remaining features, or one or more of the remaining features may be electrically coupled to the processor die through known connections and interfaces.
- a logic circuit that performs a level-conversion function in accordance with any of the aforementioned embodiments may be included in any one or more blocks shown in FIG. 6 .
- any of the blocks in FIG. 8 are operating at a different voltage, then a voltage conversion may be performed in accordance with any of the embodiments of the present invention described herein.
- these embodiments may be used to interface between the planes.
- the embodiments described herein are not tied to an specific “logic” block.
- One example could be a core-to-cache interface in a microprocessor, where the core and cache operate at different voltages. In this case, level-converting flip-flops can be used at the interface.
- the master and slave stages in the FIG. 5 are both shown to be driven by a same supply voltage.
- these stages may be driven different supply voltages, both of which may be different from the supply voltage used to drive the initial inverter 95 .
- a three-level voltage conversion may be performed.
- the supply voltage driving inverter 95 and the slave stage may be the same and the supply voltage driving the master stage may be different.
- Other variations are also possible.
- the flip-flops include an embedded level-conversion function that reduces the penalty of the level conversion compared with other designs which have been proposed.
- the disclosed embodiments also provide an easy way to interface the different voltage supply domains.
- These embodiments may also be implemented in a manner which allows the flip-flops to operate with low power and high speed while simultaneously performing the level-converting function.
- the embodiments may be implemented in a manner which avoids static current due to a low V cc signal interfacing incorrectly, which is a problem resident in many other designs. While these benefits are desirable for some applications they are not to be limiting of the invention, i.e., the embodiments of the present invention may be implemented in a manner which does not achieve one or more of these benefits.
- any reference in this specification to an “embodiment” means that a particular feature, structure, or characteristic described in connection with the embodiment is included in at least one embodiment of the invention.
- the appearances of such phrases in various places in the specification are not necessarily all referring to the same embodiment.
Abstract
Description
- This invention generally relates in one or more of its embodiments to electronic circuits which perform a level-conversion function embedded in a sequencing element.
- Many circuit applications require a voltage-level conversion to be performed for proper operation. In conventional designs, the conversion function is performed in an inefficient manner, often consuming precious chip area and increasing the processing overhead of the host system. Improvements are therefore needed in order to meet, for example, the rising demand for chip miniaturization and ever faster processing speeds.
-
FIG. 1 is a diagram showing a level-conversion function that may be performed in accordance with at least one embodiment of the present invention. -
FIG. 2 is a diagram showing a logic circuit which performs a level-conversion function in accordance with one embodiment of the present invention. -
FIG. 3A is a diagram showing that the keeper circuit in the master stage ofFIG. 2 may be formed from one or more clocked inverters,FIG. 3B is a diagram showing a transistor implementation of the inverter ofFIG. 3A ,FIG. 3C is a diagram showing that the keeper circuit in the slave stage ofFIG. 2 may be formed from clocked inverters, andFIG. 3D is a diagram showing another circuit arrangement for the slave-stage keeper circuit. -
FIG. 4 is a flow chart showing functional blocks included in a method for performing a level conversion in accordance with one embodiment of the present invention. -
FIG. 5 is a diagram showing a logic circuit which performs a level-conversion function in accordance with another embodiment of the present invention. -
FIG. 6A shows an alternative arrangement for the master-stage keeper circuit, andFIG. 6B shows an alternative implementation for one or more of the inverters which form the slave-stage keeper circuit. -
FIG. 7 is a flow chart showing functional blocks included in a method for performing a level conversion in accordance with another embodiment of the present invention. -
FIG. 8 is a diagram showing a processing system which may include a logic circuit and/or perform a level-conversion function in accordance with one or more embodiments of the present invention. -
FIG. 1 shows a logic circuit which performs a level-conversion function in accordance with one embodiment of the present invention. In this embodiment, a first voltage signal 1 is converted into asecond voltage signal 2. The first voltage signal (Vcc1) is shown as having a low voltage and the second voltage signal (Vcc2) is shown as having a high voltage. In other embodiments, the opposite conversion may be performed, e.g., a high-voltage signal may be converted into a low-voltage signal. To perform these functions, level-conversion circuitry is preferably embedded in the logic circuit in order to reduce the size and overhead of the host system while simultaneously increasing its functionality and throughput. - Also, in
FIG. 1 , it is noted that the logic circuit is shown to be a D flip-flop driven by aclock signal 3 which depending on the embodiment may be the same as or generated based on either or only one of Vcc1 or Vcc2. In other embodiments, the clock signal may be different and/or other types of flip-flops, latches, and/or logic circuits may be adapted to perform the level-conversion function of the one or more embodiments of the present invention described herein. Moreover, the type and/or magnitude of the level conversion to be performed may be adapted to meet the requirements of an intended application. -
FIG. 2 shows a circuit diagram of a D flip-flop having an embedded level-conversion circuit in accordance with one embodiment of the present invention. This flip-flop includes a master stage 10 and aslave stage 20 operating in different voltage supply domains. In this embodiment, the master stage operates in a first supply domain and includes a pass gate 11, an NMOS pull-down transistor 12, akeeper circuit 13, and an NMOS pull-down transistor 14. The keeper circuit is preferably formed from twocross-coupled inverters transistor 12 and the feedback node is coupled to the gate oftransistor 14. - The slave stage operates in a second supply domain and includes
transistors 21 and 22 and akeeper circuit 23. The keeper circuit may be formed from twocross-coupled inverters 24 and 25 and includes a storage node SN2 for storing the same logical value stored at node SN1. Feedback node FN2 stores the complement of the value stored at SN2. (The logical values stored at SN1 and SN2 are the same (e.g., logical 0 or 1), but the actual voltages corresponding to these values may differ by virtue of the different power supplies driving the master and slave stages. The same may be true for the logical values stored at FN1 and FN2.) Storage node SN2 is coupled to a terminal (drain or source) of transistor 21 and feedback node FN2 is coupled to a terminal (drain or source) oftransistor 22. The gates of these transistors are coupled to a clock signal used to alternately activate the master and slave stages. - The master and slave stages operate in different domains by virtue of the different power supplies coupled to the keeper circuits. In this embodiment, a low voltage supply is coupled to the inverters in
keeper circuit 13 and a high voltage supply is coupled to the inverters inkeeper circuit 23; that is Vcc1=Vccl and Vcc2=Vcch. The master stage thus operates at low voltage and the slave stage performs the level conversion. Put differently, the different supplies effect the conversion of the low-voltage data value stored in the master stage keeper circuit into a high-voltage data value stored in the slave stage keeper circuit. The values of Vccl and Vcch therefore control the level-conversion that takes place at the output of the flip-flop. (In an alternative embodiment, the opposite level-conversion may be performed). The low and high voltage supplies are also preferably coupled toinput inverter 16 andoutput inverter 36. - In an exemplary application, for the D input Vcc1 may be less than, equal to, or greater than Vcc2 and clock signal clk can be equal to or operate based on either Vcc1 or Vcc2. (In the
FIG. 4 embodiment, the clock signal may only operate on Vcc2). As for the keeper circuits, the logical values stored inkeeper circuits circuit 13 may be equal to or based on Vcc1 and a logical 1 incircuit 23 may be equal to or based on Vcc2. InFIG. 2 , clk and clkd are logically the same value but “d” is added to indicate a delayed clock signal. - In operation, clock signal clk activates the master and slave stages of the flip-flop at different times. In this embodiment, the clock signal is passed through two
inverters - When clk transitions from a high value to a low value, a high value (e.g., logical 1) is output from
inverter 41 and a low value is output frominverter 42. These values activate pass gate 11, thereby causing the data value from the D input to pass to storage node SN1 of thekeeper circuit 13. Becauseinverter 16 is driven by Vccl, the data value transferred through pass gate 11 operates at or based on Vccl. The voltage supply used to driveinverters - Before passing to storage node SN1, the data value is inverted by
inverter 15. The complement of the SN1 value is stored at FN1. The values stored at SN1 and FN1 are in the low-voltage domain as defined by voltage supply Vccl, which not only drivesinverter 16 but which also preferably drives the inverters of the keeper circuit. The low value of the clk turns offtransistors 21 and 22, thereby ensuring that the voltages at SN1 and FN1 are maintained until a subsequent clock signal transition takes place. - During this operation,
transistor 12 couples the value stored at SN1 to the slave stage. So when transistor 21 opens, if the value is 1 then SN2 will discharge. Conversely, if the value is 0 at SN1, when 21 opens nothing will happen to SN2, but FN2 will discharge sincetransistor 22 will be on and FN1 would have been 1 turning ontransistor 14. - When clk transitions from a low value to a high value, a low value (e.g., logical 0) is output from
inverter 41 and a high value frominverter 42. This de-activates the pass gate, thereby preventing the D input from passing into the master stage keeper circuit at this time. At the same time, a high value of clk turns ontransistors 21 and 22, thereby causing the data values stored in the master-stage keeper to pass to the slave-stage keeper circuit, e.g., the logical value at SN1 passes to SN2 and the logical value at FN1 passes to FN2. The value at FN2 then passes to the Q output of the flip-flop throughinverter 36, which is also preferably driven by the slave-stage power supply. - The slave stage of the flip-flop therefore performs the internal level-conversion function of the flip-flop, converting the input data value from the low-voltage domain (Vccl) to the high-voltage domain (Vcch) at the output. This is accomplished by driving at least the keeper circuit of the slave stage with a higher voltage supply than is used to drive the master stage keeper. Also, the
slave stage inverter 36 is preferably driven using the higher domain voltage and the clock signal clk may be the same as or different from the high-Vcc Q output. - While the
FIG. 2 embodiment performs a level conversion from a low voltage supply domain to a high voltage supply domain, the opposite level conversion can be performed by drivinginverters inverters -
FIGS. 3A-3D show alternative arrangements for the keeper circuits.FIG. 3A shows thatinverter 16 may be a clocked inverter which store and output data based on clock signal clk and its complement {overscore (clk)}. -
FIG. 3B shows one type of transistor implementation of the inverter ofFIG. 3A . This implementation includes twoPMOS transistors transistors transistors transistors transistors -
FIG. 3C shows thatkeeper circuit 20 may be formed as an arrangement of clocked inverters, e.g.,inverters -
FIG. 3D shows another implementation ofkeeper circuit 20. This implementation includes a cross-coupled arrangement of PMOS transistors 47 through 50. The gates oftransistors FIGS. 3A-3D may be incorporated intoFIG. 2 for purposes of storing data therein. -
FIG. 4 is a flow chart showing functional blocks included in a method for performing a level conversion in accordance with one embodiment of the present invention. The method may be performed by the circuit ofFIG. 2 or by another circuit. The method includes storing an input data value in a first stage of a flip-flop circuit (Block 51) and then driving the first stage with a first supply voltage (Block 52). A second stage of the flip-flop circuit is then activated (Block 53), and then the value stored in the first stage of the flip-flop circuit is transferred for storage in the second stage (Block 54). The second stage is then driven with a second supply voltage different from the first supply voltage, thereby achieving the level conversion. (Block 55). The value stored in the second stage is then used as a basis for forming the output of the flip-flop. (Block 56). Preferably, the first and second stages are alternately activated by different states of a clock signal. -
FIG. 5 shows a circuit diagram of a D flip-flop having an embedded level-conversion circuit in accordance with another embodiment of the present invention. This flip-flop includes amaster stage 100 and aslave stage 110 which operate in a same voltage supply domain. An inverter 95 coupled to the D input is driven by a different supply voltage. InFIG. 4 , the master and slave stages are shown as being driven by a high voltage supply and inverter 95 by a low voltage supply. This embodiment may prove to be preferable for some applications, since by driving the master stage with a higher voltage supply will make the flip-flop less susceptible to noise. - The master stage includes a pull-down clocked NMOS transistor 111, a pull-down clocked
NMOS transistor 112, and akeeper circuit 113 formed fromcross-coupled inverters transistor 112 is coupled to the feedback node FN1′. Inverter 95, which inverts the input data value for storage at the feedback node FN1′, is coupled between the gates of twotransistors Transistors transistors 111 and 112 turn on, either 101 or 102 will pass a 0 to either SN1′ or FN1′ respectively. If SN1′ is 0, theninverter 114 will pull FN1′ to 1. - The slave stage includes a
keeper circuit 121 having a storage node SN2′ and feedback node FN2′ disposed betweencross-coupled inverters pass gate 130, which is operated based on an input clock signal clk2. In this embodiment, only one clocked inverter is used in order to reduce the clock power of the flip-flop. Also, in this embodiment the master and slave stages operate in a same voltage domain (e.g., Vcch) and only the input inverter 95 is driven by a different voltage supply. Since all stage nodes are driven at a high-voltage supply Vcch, this embodiment may prove to be more robust to noise than the previous embodiment. In all embodiments described herein, the feedback inverters in the keeper circuits may be sized to be weak or they can be fully interrupted based on the clock signal. - In operation, clock signal clk′ activates the master and slave stages of the flip-flop at different times. When clk′ transitions from a high value to a low value, a high value is output from
inverter 140, thereby turning ontransistors 111 and 112. As a result, the data value at the D input and its complement (generated by the output of inverter 95) respectively pass to storage and feedback nodes SN1′ and FN1′ of the master keeper circuit. Because the clock is low, the pass gate is not activated and thus the value stored at FN1′ does not pass to the slave stage. - When clock signal clk′ transitions from a low to a high value, a low value is output from
inverter 140 to isolate the master stage from the D input. At the same time,pass gate 130 is activated to pass the logical value stored at node FN1′ to the slave-stage keeper circuit through inverter 120. Preferably, bothinverters 120 and 140 are driven by the same supply voltage, which, for example, may be the same supply voltage used to drive the master and slave stages, e.g., Vcch. In an alternative embodiment, these inverters may be driven by different supply voltages, at least one of which may differ from the master and slave stage supplies. - In the slave stage, the logical value output from inverter 120 is stored at node SN2′ of
keeper circuit 121 and the complement of this value is stored at node FN2′. The logical value stored at node SN2′ is passed through an inverter 150 to form the Q output of the flip-flop. To maintain the level conversion, inverter 150 is driven by master- and slave-stage supply voltage, e.g., Vcch. Clock signal clk2 may equal or be based on Vcc2 inFIG. 1 , which in this case may be the high-voltage power supply Vcch. -
FIG. 6A shows an alternative arrangement for the master-stage keeper circuit. In this version of the keeper,inverters -
FIG. 6B shows an alternative implementation for one or more of the inverters which form the slave-stage keeper circuit. In this implementation,inverter 125 is a clocked inverter and the other inverter in this keeper may have a similar arrangement.Inverter 125 functions as an inverter when clk2 is low and {overscore (clk)} 2 is high. When clk2 is high and {overscore (clk)} 2 is low, the inverter is tri-stated allowing pass gate 30 to write into SN2′ without contention. Basically, when this pass gate is closed,inverter 125 functions like an inverter and when 20 is open the inverter is tri-stated. -
FIG. 7 is a flow chart showing functional blocks included in a method for performing a level conversion in accordance with another embodiment of the present invention. The method may be performed by the circuit ofFIG. 5 or by another circuit. The method includes passing an input data value through an inverter (Block 151) and then driving the inverter with a first supply voltage (Block 152). The inverted value is then transferred for storage into a first stage of a flip-flop circuit. (Block 153). A second stage of the flip-flop circuit is then activated (Block 154), and the value stored in the first stage is transferred through another inverter for storage in the second stage (Block 155). The first and second stages are driven with at least a second supply voltage. Depending on the application, the first and second stages may be driven by different supply voltages, both of which are different from the first supply voltage. The value stored in the second stage is then used as a basis for forming the output of the flip-flop. (Block 156). Depending on the type of flip-flop, the second stage output value may be inverted. Preferably, the first and second stages are alternately activated by different states of a clock signal. - The level conversion performed by foregoing embodiments are illustratively described as being included in flip-flop circuits. The use of flip-flop circuits, while preferable, is by no means a necessity. Rather, the level conversions described herein may be incorporated in a variety of other logic circuits including but not limited to latches, flip-flops, and mux latches.
- Also, the logic circuits may themselves be included in a variety of higher-level circuits or systems, one example of which includes a microprocessor system operating in two or more voltage supply (Vcc) domains. In one common design, one domain operates based on a low voltage supply (Vccl) and another domain operates based on a high voltage supply (Vcch). Or, in an adaptive voltage environment, the first voltage domain can be less than or greater than the second voltage domain, or the domains may be equal, at any time during operation. One or more embodiments of the present invention may be used to perform a level-conversion between these domains. Put differently, a logic circuit in accordance with any one or more of the embodiments previously described may be used, for example, to latch or store a data value while simultaneously serving as a level-conversion interface between low-Vcc and high-Vcc domains. Other higher-level systems which may include a logic circuit in accordance with any of the embodiments described herein include but are not limited to multi-core processors, system-on-chip, core-to-cache interface, and multi-voltage ASICS.
-
FIG. 8 is a diagram of a processing system which includes any one or more embodiments of the present invention. The processing system includes aprocessor 210, apower supply 220, and amemory 230 which, for example, may be a random-access memory. The processor may include anarithmetic logic unit 212 and aninternal cache 214. The system may also include agraphical interface 240, achipset 250, acache 260 and anetwork interface 270. The processor may be a microprocessor or any other type of processor. Also, the processor may be included on a chip die with all or any combination of the remaining features, or one or more of the remaining features may be electrically coupled to the processor die through known connections and interfaces. A logic circuit that performs a level-conversion function in accordance with any of the aforementioned embodiments may be included in any one or more blocks shown inFIG. 6 . - For example, if any of the blocks in
FIG. 8 are operating at a different voltage, then a voltage conversion may be performed in accordance with any of the embodiments of the present invention described herein. Thus, if any of the blocks have different voltage planes, these embodiments may be used to interface between the planes. The embodiments described herein are not tied to an specific “logic” block. One example could be a core-to-cache interface in a microprocessor, where the core and cache operate at different voltages. In this case, level-converting flip-flops can be used at the interface. - The embodiments previously discussed may be modified in various ways. For example, the master and slave stages in the
FIG. 5 are both shown to be driven by a same supply voltage. In an alternative embodiment, these stages may be driven different supply voltages, both of which may be different from the supply voltage used to drive the initial inverter 95. In this case, a three-level voltage conversion may be performed. As another alternative, the supply voltage driving inverter 95 and the slave stage may be the same and the supply voltage driving the master stage may be different. Other variations are also possible. - In the foregoing embodiments, the flip-flops include an embedded level-conversion function that reduces the penalty of the level conversion compared with other designs which have been proposed. Through the flip-flops, the disclosed embodiments also provide an easy way to interface the different voltage supply domains. These embodiments may also be implemented in a manner which allows the flip-flops to operate with low power and high speed while simultaneously performing the level-converting function. The embodiments may be implemented in a manner which avoids static current due to a low Vcc signal interfacing incorrectly, which is a problem resident in many other designs. While these benefits are desirable for some applications they are not to be limiting of the invention, i.e., the embodiments of the present invention may be implemented in a manner which does not achieve one or more of these benefits.
- Any reference in this specification to an “embodiment” means that a particular feature, structure, or characteristic described in connection with the embodiment is included in at least one embodiment of the invention. The appearances of such phrases in various places in the specification are not necessarily all referring to the same embodiment. Further, when a particular feature, structure, or characteristic is described in connection with any embodiment, it is submitted that it is within the purview of one skilled in the art to affect such feature, structure, or characteristic in connection with other ones of the embodiments.
- Furthermore, for ease of understanding, certain functional blocks may have been delineated as separate blocks; however, these separately delineated blocks should not necessarily be construed as being in the order in which they are discussed or otherwise presented herein. For example, some blocks may be able to be performed in an alternative ordering, simultaneously, etc.
- Although the present invention has been described herein with reference to a number of illustrative embodiments, it should be understood that numerous other modifications and embodiments can be devised by those skilled in the art that will fall within the spirit and scope of the principles of this invention. More particularly, reasonable variations and modifications are possible in the component parts and/or arrangements of the subject combination arrangement within the scope of the foregoing disclosure, the drawings and the appended claims without departing from the spirit of the invention. In addition to variations and modifications in the component parts and/or arrangements, alternative uses will also be apparent to those skilled in the art.
Claims (28)
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US10/924,906 US7132856B2 (en) | 2004-08-25 | 2004-08-25 | Hybrid CVSL pass-gate level-converting sequential circuit for multi-Vcc microprocessors |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US10/924,906 US7132856B2 (en) | 2004-08-25 | 2004-08-25 | Hybrid CVSL pass-gate level-converting sequential circuit for multi-Vcc microprocessors |
Publications (2)
Publication Number | Publication Date |
---|---|
US20060044013A1 true US20060044013A1 (en) | 2006-03-02 |
US7132856B2 US7132856B2 (en) | 2006-11-07 |
Family
ID=35942212
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US10/924,906 Expired - Fee Related US7132856B2 (en) | 2004-08-25 | 2004-08-25 | Hybrid CVSL pass-gate level-converting sequential circuit for multi-Vcc microprocessors |
Country Status (1)
Country | Link |
---|---|
US (1) | US7132856B2 (en) |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20140240002A1 (en) * | 2013-02-22 | 2014-08-28 | Hideep Inc | Voltage level converor and rf switching driver using the same |
EP2869467A1 (en) * | 2013-11-01 | 2015-05-06 | Nxp B.V. | Latch circuit |
US20180069538A1 (en) * | 2016-09-08 | 2018-03-08 | Intel Corporation | Low clock supply voltage interruptible sequential |
Families Citing this family (13)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
KR100833179B1 (en) * | 2006-02-15 | 2008-05-28 | 삼성전자주식회사 | Level converting flip-flop and pulse generator for clusterd voltage scaling |
US7622976B2 (en) * | 2006-07-06 | 2009-11-24 | Stc.Unm | Sequential circuit design for radiation hardened multiple voltage integrated circuits |
JP2008199153A (en) * | 2007-02-09 | 2008-08-28 | Matsushita Electric Ind Co Ltd | Level shifter |
JP4289410B2 (en) * | 2007-03-12 | 2009-07-01 | セイコーエプソン株式会社 | Level shift circuit, electro-optical device, and level shift method |
US7956662B2 (en) * | 2009-01-16 | 2011-06-07 | Freescale Semiconductor, Inc. | Flip-flop circuit with internal level shifter |
US7855575B1 (en) | 2009-09-25 | 2010-12-21 | Intel Corporation | Wide voltage range level shifter with symmetrical switching |
US8362803B2 (en) * | 2011-02-18 | 2013-01-29 | Lsi Corporation | Mode latching buffer circuit |
US8378728B1 (en) * | 2012-06-03 | 2013-02-19 | Freescale Semiconductor, Inc. | Level shifting flip-flop |
US9385722B2 (en) | 2014-11-25 | 2016-07-05 | Intel Corporation | Voltage level shifter circuit |
US9432002B2 (en) * | 2014-12-11 | 2016-08-30 | Freescale Semiconductor, Inc. | High-speed voltage level shifter circuit |
US9553584B2 (en) * | 2014-12-23 | 2017-01-24 | International Business Machines Corporation | Level-shifting latch |
US9590599B1 (en) * | 2015-10-09 | 2017-03-07 | Nxp B.V. | Clock distribution for flip-flop circuits |
US20180091150A1 (en) * | 2016-09-27 | 2018-03-29 | Intel Corporation | Fused voltage level shifting latch |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5612632A (en) * | 1994-11-29 | 1997-03-18 | Texas Instruments Incorporated | High speed flip-flop for gate array |
US6429711B1 (en) * | 2000-06-30 | 2002-08-06 | Intel Corporation | Stack-based impulse flip-flop with stack node pre-charge and stack node pre-discharge |
US6535433B2 (en) * | 1999-02-22 | 2003-03-18 | Mitsubishi Denki Kabushiki Kaisha | Semiconductor logic circuit device of low current consumption |
US6781411B2 (en) * | 2002-01-29 | 2004-08-24 | Texas Instruments Incorporated | Flip flop with reduced leakage current |
US6809554B2 (en) * | 2002-09-20 | 2004-10-26 | Kabushiki Kaisha Toshiba | Semiconductor integrated circuit having a voltage conversion circuit |
US20050285624A1 (en) * | 2004-06-29 | 2005-12-29 | Hsu Steven K | Hybrid pass gate level converting dual supply sequential circuit |
-
2004
- 2004-08-25 US US10/924,906 patent/US7132856B2/en not_active Expired - Fee Related
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5612632A (en) * | 1994-11-29 | 1997-03-18 | Texas Instruments Incorporated | High speed flip-flop for gate array |
US6535433B2 (en) * | 1999-02-22 | 2003-03-18 | Mitsubishi Denki Kabushiki Kaisha | Semiconductor logic circuit device of low current consumption |
US6429711B1 (en) * | 2000-06-30 | 2002-08-06 | Intel Corporation | Stack-based impulse flip-flop with stack node pre-charge and stack node pre-discharge |
US6781411B2 (en) * | 2002-01-29 | 2004-08-24 | Texas Instruments Incorporated | Flip flop with reduced leakage current |
US6809554B2 (en) * | 2002-09-20 | 2004-10-26 | Kabushiki Kaisha Toshiba | Semiconductor integrated circuit having a voltage conversion circuit |
US20050285624A1 (en) * | 2004-06-29 | 2005-12-29 | Hsu Steven K | Hybrid pass gate level converting dual supply sequential circuit |
Cited By (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20140240002A1 (en) * | 2013-02-22 | 2014-08-28 | Hideep Inc | Voltage level converor and rf switching driver using the same |
US8963583B2 (en) * | 2013-02-22 | 2015-02-24 | Hideep Inc. | Voltage level converter and RF switching driver apparatus using the same |
EP2869467A1 (en) * | 2013-11-01 | 2015-05-06 | Nxp B.V. | Latch circuit |
US9490782B2 (en) | 2013-11-01 | 2016-11-08 | Nxp B.V. | Latch circuit |
US9490781B2 (en) | 2013-11-01 | 2016-11-08 | Nxp B.V. | Redundant clock transition tolerant latch circuit |
US20180069538A1 (en) * | 2016-09-08 | 2018-03-08 | Intel Corporation | Low clock supply voltage interruptible sequential |
US10418975B2 (en) * | 2016-09-08 | 2019-09-17 | Intel Corporation | Low clock supply voltage interruptible sequential |
Also Published As
Publication number | Publication date |
---|---|
US7132856B2 (en) | 2006-11-07 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US7132856B2 (en) | Hybrid CVSL pass-gate level-converting sequential circuit for multi-Vcc microprocessors | |
US9742382B2 (en) | Flip-flop for reducing dynamic power | |
US7365575B2 (en) | Gated clock logic circuit | |
US6459316B1 (en) | Flip flop circuit | |
US7301372B2 (en) | Domino logic compatible scannable flip-flop | |
JP3614125B2 (en) | CP flip-flop | |
US6188260B1 (en) | Master-slave flip-flop and method | |
US6720813B1 (en) | Dual edge-triggered flip-flop design with asynchronous programmable reset | |
US6864732B2 (en) | Flip-flop circuit with reduced power consumption | |
US6864733B2 (en) | Data-enabled static flip-flop circuit with no extra forward-path delay penalty | |
US10033356B2 (en) | Reduced power set-reset latch based flip-flop | |
JPH11186882A (en) | D flip-flop | |
US10903824B2 (en) | Pulsed level shifter circuitry | |
EP1732227B1 (en) | N-Domino output latch | |
US20160269002A1 (en) | Low clocking power flip-flop | |
EP1693964B1 (en) | P-Domino output latch | |
US6707318B2 (en) | Low power entry latch to interface static logic with dynamic logic | |
US7449924B2 (en) | Latch-based serial port output buffer | |
EP1732226A2 (en) | P-domino register | |
KR20230154198A (en) | Circuit units, logic circuits, processors, and computing devices | |
US6919737B2 (en) | Voltage-level converter | |
US6407604B1 (en) | Register and latch circuits | |
JP6386130B2 (en) | Flip-flop to reduce dynamic power | |
JPH1093397A (en) | D type flip-flop | |
US20090201063A1 (en) | Dynamic semiconductor device |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
AS | Assignment |
Owner name: INTEL CORPORATION, CALIFORNIA Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNORS:HSU, STEVEN K.;KRISHNAMURTHY, RAM K.;REEL/FRAME:015737/0209 Effective date: 20040818 |
|
FPAY | Fee payment |
Year of fee payment: 4 |
|
FPAY | Fee payment |
Year of fee payment: 8 |
|
FEPP | Fee payment procedure |
Free format text: MAINTENANCE FEE REMINDER MAILED (ORIGINAL EVENT CODE: REM.) |
|
LAPS | Lapse for failure to pay maintenance fees |
Free format text: PATENT EXPIRED FOR FAILURE TO PAY MAINTENANCE FEES (ORIGINAL EVENT CODE: EXP.); ENTITY STATUS OF PATENT OWNER: LARGE ENTITY |
|
STCH | Information on status: patent discontinuation |
Free format text: PATENT EXPIRED DUE TO NONPAYMENT OF MAINTENANCE FEES UNDER 37 CFR 1.362 |
|
FP | Lapsed due to failure to pay maintenance fee |
Effective date: 20181107 |